Add annotation transfers

[dosumis]

https://static-content.springer.com/esm/art%3A10.1038%2Fs41586-023-06812-z/MediaObjects/41586_2023_6812_MOESM8_ESM.xlsx

Cluster table includes:

column	description
CTX.cluster_id	The best matching cluster id in the previous published CTX_HPF taxonomy (https://portal.brain-map.org/atlases-and-data/rnaseq/mouse-whole-cortex-and-hippocampus-10x)
CTX.cluster_label	The best matching cluster label in the previous published CTX_HPF taxonomy
CTX.supertype_id	The best matching supertype id in the previous published CTX_HPF taxonomy
CTX.supertype_label	The best matching supertype label in the previous published CTX_HPF taxonomy
CTX.subclass_id	The best matching subclass id in the previous published CTX_HPF taxonomy
CTX.subclass_id	The best matching subclass label in the previous published CTX_HPF taxonomy
CTX.neighborhood_id	The best matching neighborhood id in the previous published CTX_HPF taxonomy
CTX.neighborhood_label	The best matching neighborhood label in the previous published CTX_HPF taxonomy
CTX.size	The number of cells included in the previous published CTX_HPF taxonomy

e.g.

CTX.cluster_id	CTX.cluster_label	CTX.supertype_id	CTX.supertype_label	CTX.subclass_id	CTX.subclass_id	CTX.neighborhood_id
238	238_Car3_2	45	Car3	21	Car3	4
238	238_Car3_3	45	Car3	21	Car3	4

We need a taxonomy for https://portal.brain-map.org/atlases-and-data/rnaseq/mouse-whole-cortex-and-hippocampus-10x

This corresponds to https://cellxgene.cziscience.com/collections/e3aa612b-0d7d-4d3f-bbea-b8972a74dd4b. We should be able to build a taxonomy directly from the 10x dataset (https://cellxgene.cziscience.com/e/d7291f04-fbbb-4d65-990a-f01fa44e915b.cxg/) using CAS-tools. But we should also check whether an existing AIT taxonomy is available.
@patrick-lloyd-ray -can you check this?

Once this is in place, we can then translate then use the table to generate annotation transfer entries in this taxonomy

TODO: Check paper for annotation transfer method details (not found so far, but may be buried?)

[dosumis]

Options for building taxonomy:

CAS tools from https://cellxgene.cziscience.com/e/d7291f04-fbbb-4d65-990a-f01fa44e915b.cxg/

But also in JSON here: https://portal.brain-map.org/atlases-and-data/rnaseq/mouse-whole-cortex-and-hippocampus-10x @hkir-dev has python that can work with the JSON. Might be additional info in their we can use. Update: JSON has brain regions as well as cell types, so might not look like the others we've processed. Working with this is therefore lower priority.

[patrick-lloyd-ray]

Yeah, only the JSON that you linked above.

[dosumis]

@ubyndr or @hkir-dev could you create CAS for this using CAS https://portal.brain-map.org/atlases-and-data/rnaseq/mouse-whole-cortex-and-hippocampus-10x & generate a TDT repo?

I think the naming should be based on the taxonomy JSON https://idk-etl-prod-download-bucket.s3.amazonaws.com/aibs_mouse_ctx-hpf_10x/dend.json

This can be used to name the CAS and set cell_set_accessions.

We could also potentially mine synonyms and refs from it:

e.g.

 {
                                                      "label": "CTX-HPF 007",
                                              
                                                      "cell_set_accession": "CS202106160_7",
                                                      "original_label": "7_Lamp5 Lhx6",
                                                      "cell_set_designation": "CTX-HPF 007",
                                                      "cell_set_alias": "7_Lamp5 Lhx6",
                                                      "cell_set_structure": "Cortical plate; Hippocampal formation",
                                                      "cell_set_ontology_tag": "none; none",
                                                      "cell_set_alias_assignee": "Zizhen Yao",
                                                      "cell_set_alias_citation": "https://doi.org/10.1016/j.cell.2021.04.021",
                                                      "taxonomy_id": "CCN202106160",       
}